Has value investing gone out of style?
Is Value dead? Or do the Russell Index family and the Fama-French1 factor enthusiasts need to re-evaluate price-to-book ratio as the primary metric used in defining value stocks?
Taking a step back for a moment, there is no “universal” definition of what value truly is within the investment industry. Practitioners seem to “know-it-when-they-see-it”, but to get there they could be using almost any combination of fundamental statistics, some of which could include:
- Price-to-Earnings (P/E) ratio—a lower P/E ratio would denote a more value-oriented strategy.
- Price-to-Book ratio—a lower price-to-book ratio would denote a more value-oriented strategy.
- Price-cash-flow ratio—a lower price-to-cash flow ratio would denote a more value-oriented strategy.
- Price-to-sales ratio—a lower price-to-sales ratio would denote a more value-oriented strategy.
- Dividend Yield—the prior four metrics position the ratio as the price relative to a fundamental. Any “yield” measure flips the ratio and positions the fundamental relative to the price. Since a higher fundamental (in this case, dividend) measured against a lower price would be indicative of a more value-oriented strategy, a higher dividend yield would be more “value” than a lower dividend yield.
Each ratio has its own set of pros and cons, especially since certain types of companies may be biased toward running their businesses in certain ways. For example, Financials firms tend to typically have lower price to book ratios (and high weightings in value approaches that use this statistic) than many firms in other industries. The issue is that a high book value of equity based on how a particular industry does its accounting practices may bias towards a lower price to book ratio, as the ratio doesn’t go down solely because the price has fallen. Many strategies use a mixture of valuation ratios to denote value so as to mitigate the risks of such biases.
Getting back to our unique findings on price to book: the decade ending September 2018 was the single worst decade in history for the price-to-book ratio as a factor sort of the market.
Figure 1: Rolling 10-Year return differentials: Cheap versus expensive Price-to-Book (P/B)
Source: Fama-French Data Library, 30 June 1936 to 30 November 2018. “Expensive” based on lowest 30% of companies ranked based on Book-to-Price and “Cheap” based on the highest 30%. Returns based on available data for companies included in the Center for Research in Security Prices (CRSP) database and listed on the New York Stock Exchange, American Stock Exchange or NASDAQ stock exchange. Portfolios are market-capitalisation weighted and rebalanced annually at the end of June.
Historical performance is not an indication of future performance and any investments may go down in value. You cannot invest directly in an index.
Deteriorating power?
In their seminal factor paper, Fama and French wrote: “there is no evidence that its explanatory power deteriorates through time.”2
Perhaps they will consider the new evidence.
In 616 of the 679 10-year periods3 from 1926 through the end of 1992, the cheap stocks outperformed the expensive stocks with the median performance advantage greater than 500 basis points (bps) per year.
But recent returns have been trending down as we saw earlier with the recent single worst decade for low price-to-book stocks compared to high price-to-book stocks.
This is not just an academic exercise on the limits to valuation-oriented investing.
The Russell family of indexes established in 1979—ahead of the Fama-French research—utilizes the price-to-book ratio as the primary determinant of what defines a value stock. It now has been 40-years since the creation of that index family and the Russell 1000 Value Index has out-performed the Russell 1000 Growth Index.
But consistent with the recent trending down in outperformance, over the past 30 years, the Russell 1000 Value Index has lagged the Russell 1000 Growth by 21 bps annualized, with more recent underperformance significantly greater.4
What caused recent under performance of value?
Primarily sector tilts caused the lagging performance for Russell Value. Price to book sorting of the market as a value strategy results in chronic sector tilts, particularly towards Financial stocks and away from Technology stocks.
On average, the Russell 1000 Value Index has been overweight Financials by 11%, and underweight Information Technology by 10% over last decade. These bets are persistent and likely will continue to for some time.
Figure 2: Russell 1000 Value Index versus Russell 1000 Index Average Active Weights: Trailing 10 years
Source: WisdomTree, FactSet.
Historical performance is not an indication of future performance and any investments may go down in value. You cannot invest directly in an index.
A sector attribution of the Russell 1000 Index performance relative to the Russell 1000 Value Index over the past decade can help dissect this performance impact.
Of the 210 bps of outperformance for the Russell 1000 Index over the Russell 1000 Value, 83 bps came from the Information Technology sector alone.
Figure 3: Trailing 10-year attribution of Russell 1000 Index vs. Russell 1000 Value Index: Information Technology and Financials
Source: WisdomTree, FactSet, 31 December 2008 to 31 December 2018.
Historical performance is not an indication of future performance and any investments may go down in value. You cannot invest directly in an index.
A number of market commentators believe the epic run of growth stock out-performance will turn and that value is destined to turn around.
This may be true, but clearly the most important ingredient for that turn would be Financials out-performing Technology. How confident are you in that outcome?
WisdomTree will be coming back to this topic more in 2019, particularly focusing on new value sensitive index strategies that do not contain persistent sector biases.
Sources
1 Refers to the Fama–French three-factor model designed by Eugene Fama and Kenneth French to describe stock returns.
2 Fama French Research, 1992.
3 Refers to the 679 10-year periods (i.e. June 1926 to 1936; July 1926 to 1936 and so on) from 30 June 1926 to 31 December 1992, rolled forward on a monthly basis.
4 Zephyr StyleAdvisor, 31 December 1988 to 31 December 2018.
Related blogs
+ Growth stocks: Will they clear the hurdle?